USPTO Trademark Checker: Bulk Phrase Risk Check for Print on Demand Sellers

Paste a list of slogans, product names or brand ideas and get one clear answer per phrase: clear, low, medium or high trademark risk in the US, with the live marks behind the answer. Built for print-on-demand, Etsy, Amazon Merch, Redbubble and Shopify sellers who screen hundreds of phrases before they upload a design.

The Actor searches the public USPTO trademark database (the same data as the official search page), compares every phrase with the registered and pending marks, and rates the risk for your product classes (for example class 25 for T-shirts and hats, class 21 for mugs). It is a fast first screen. It is not legal advice and not a clearance opinion.

• $0.05 per phrase checked. Phrases that could not be checked are free. No monthly fee, no start fee beyond the Apify platform default.
• Up to 1,000 phrases per run. Duplicates are checked once.
• No login, no API key, no proxy.
• Export as JSON, CSV, Excel or call it through the Apify API.

What you get for every phrase

Field	Meaning
verdict	clear, low, medium, high, or error
verdictReason	One sentence, for example "3 live exact match(es) in your classes"
liveExactMatches, liveContainingMatches, liveSimilarMatches	How many live marks match the phrase exactly, contain it (or sit inside it), or look similar
topMatchWordmark, topMatchOwner, topMatchStatus, topMatchSerialNumber	The most serious conflict, flat fields for spreadsheets
matches[]	Conflicting marks, most serious first: serial and registration number, wording, match type, status, classes, whether it sits in your class, owner, filing and registration date, goods and services, link to the USPTO case page

How the verdict is decided

Verdict	Rule
high	A live (registered or pending) mark with the same wording in one of your classes
medium	A live mark that contains your phrase or sits inside it in your classes, or a live exact match in another class
low	Only similar spellings in your classes, related marks in other classes, or dead marks
clear	No live matching mark found

If you leave the classes empty, every class counts the same.

Example input

{
  "phrases": ["Just do it", "Cat Dad Life", "Zxqvk Plumbo Wrenly"],
  "classes": ["25"],
  "matchMode": "exact-and-similar",
  "maxMatchesPerPhrase": 5
}

Example output (shortened)

{
  "phrase": "Just do it",
  "verdict": "high",
  "verdictReason": "3 live exact match(es) in your classes",
  "liveExactMatches": 12,
  "topMatchWordmark": "JUST DO IT.",
  "topMatchOwner": "NIKE, INC.",
  "topMatchStatus": "registered",
  "matches": [
    { "serialNumber": "73829171", "wordmark": "JUST DO IT.", "matchType": "exact", "status": "registered", "classes": ["25"], "inRequestedClass": true, "ownerName": "NIKE, INC.", "usptoUrl": "https://tsdr.uspto.gov/#caseNumber=73829171&caseType=SERIAL_NO&searchType=statusSearch" }
  ],
  "source": "USPTO trademark search (tmsearch.uspto.gov)"
}

Input

• Phrases to check: one per line, up to 1,000.
• Your product classes: class 25 clothing, 21 mugs and housewares, 9 phone cases, 16 posters and stickers, 18 bags, 20 pillows, 24 blankets, and more. Default is 25.
• Match mode: exact and similar (default, one extra search per phrase) or exact only.
• Include dead marks: also list abandoned and cancelled marks. They never raise the verdict above low.
• Matches per phrase: 1 to 50, default 10.
• Show names of individual owners: off by default. When a mark belongs to a private person the name is withheld; company names are always shown.

Use it from code or with AI agents

curl -X POST "https://api.apify.com/v2/acts/atesen-software~trademark-phrase-risk-check/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"phrases":["Mama Bear","Cat Dad Life"],"classes":["25","21"]}'

The Actor works with the Apify MCP server, so an AI agent can screen a list of design ideas and keep only the clear and low ones. Make, Zapier and n8n can call it through the Apify integrations.

Good to know

• US marks only. The data is the USPTO register. A phrase that is clear here can still be protected in the EU, the UK or elsewhere.
• Word marks. Logos and design marks are not compared visually. Wording, spelling and word order are.
• Speed. The Actor sends about one request per second to be polite to the USPTO servers. A run of 100 phrases takes roughly five minutes.
• Freshness. Results are read live. New applications appear as soon as the USPTO publishes them.
• Not legal advice. A clear verdict means no live matching mark was found in this search. It does not guarantee you can use a phrase. Ask a trademark attorney before you build a brand on it.
• If the USPTO answers with a bot check or a block, the run stops, marks the remaining phrases as error (free) and never tries to get around it.

Data and privacy

The data comes from the public USPTO trademark records. The Actor reads marks, status, classes and owner names of companies. Owner addresses and attorney details are not collected, and names of private persons are withheld unless you switch them on.